To cite this article before publication: Zhaomin Dong et al 2021 Environ. Res. Lett. in press https://doi.org/10.1088/1748-9326/abda71
Manuscript version: Accepted Manuscript
Accepted Manuscript is “the version of the article accepted for publication including all changes made as a result of the peer review process, and which may also include the addition to the article by IOP Publishing of a header, an article ID, a cover sheet and/or an ‘Accepted Manuscript’ watermark, but excluding any other editing, typesetting or other changes made by IOP Publishing and/or its licensors” This Accepted Manuscript is © 2020 The Author(s). Published by IOP Publishing Ltd.
As the Version of Record of this article is going to be / has been published on a gold open access basis under a CC BY 3.0 licence, this Accepted Manuscript is available for reuse under a CC BY 3.0 licence immediately.
Everyone is permitted to use all or part of the original content in this article, provided that they adhere to all the terms of the licence
https://creativecommons.org/licences/by/3.0
Although reasonable endeavours have been taken to obtain all necessary permissions from third parties to include their copyrighted content within this article, their full citation and copyright line may not be present in this Accepted Manuscript version. Before using any content from this article, please refer to the Version of Record on IOPscience once published for full citation and copyright details, as permissions may be required. All third party content is fully copyright protected and is not published on a gold open access basis under a CC BY licence, unless that is specifically stated in the figure caption in the Version of Record.
View the article online for updates and enhancements.
1
1 Data-related and methodological obstacles to determining associations between
2 temperature and COVID-19 transmission
3
4 Zhaomin Dong1,2, Xiarui Fan1, Jiao Wang3, Yixin Mao3, Yueyun Luo3, and Song Tang3,4*
5
6 1 School of Space and Environment, Beihang University, Beijing 100191, China
7 2 Beijing Advanced Innovation Center for Big Data-Based Precision Medicine, Beihang
8 University, Beijing 100191, China
9 3 China CDC Key Laboratory of Environment and Population Health, National Institute of
10 Environmental Health, Chinese Center for Disease Control and Prevention, Beijing 100021,
11 China
12 4 Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing,
13 Jiangsu 211166, China
14
15 * Correspondence
16 No. 7 Panjiayuan Nanli, Chaoyang District, Beijing 100021, China. Email:
17 [email protected] (Dr. Song Tang)
18 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
2
19 Abstract
20 More and more studies have evaluated the associations between ambient temperature and
21 coronavirus disease 2019 (COVID-19). However, most of these studies were rushed to
22 completion, rendering the quality of their findings questionable. We systematically evaluated
23 70 relevant peer-reviewed studies published on or before September 21, 2020 that had been
24 implemented from community to global level. Approximately 35 of these reports indicated that
25 temperature was significantly and negatively associated with COVID-19 spread, whereas 12
26 reports demonstrated a significantly positive association. The remaining studies found no
27 association or merely a piecewise association. Correlation and regression analyses were the
28 most commonly utilized statistical models. The main shortcomings of these studies included
29 uncertainties in COVID-19 infection rate, problems with data processing for temperature,
30 inappropriate controlling for confounding parameters, weaknesses in evaluation of effect
31 modification, inadequate statistical models, short research periods, and the choices of research
32 areal units. It is our viewpoint that most studies of the identified 70 publications have had
33 significant flaws that have prevented them from providing a robust scientific basis for the
34 association between temperature and COVID-19.
35
36 Keywords: SARS-CoV-2; temperature; transmission; methodological concerns; data
37 uncertainties 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
3 38 Graphical Abstract 39 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
4
40 1. Introduction
41 The coronavirus disease 2019 (COVID-19) pandemic, which is ongoing at the time of
42 writing, has attracted increasing research interests (Gong et al. 2020). An understanding of the
43 driving factors of COVID-19 transmission is urgently needed owing to the extensive public
44 health implications (Kraemer et al. 2020). Whether warm temperatures suppress the spread of
45 COVID-19 has become a hot topic of discussion that has attracted considerable social media
46 and political attention worldwide, since preliminary laboratory studies indicated the high
47 temperature can lower the survival of COVID-19 virus (Baker et al. 2020; NAS. 2020).
48 Inputting the keywords “temperature” and “COVID-19” into the Web of Science yielded
49 hundreds of results (as of September 21, 2020), but the main findings of these publications
50 were not consistent (Fang et al. 2020; Jüni et al. 2020; Pan et al. 2020). As a large proportion
51 of this research had been conducted in a rush (Glasziou et al. 2020; Heederik et al. 2020), its
52 findings may be more likely to generate public confusion than to contribute to scientific
53 knowledge (Zeka et al. 2020). A recent study criticized all of the studies associated with
54 ambient air pollution and COVID-19 incidence and mortality, arguing that they were
55 susceptible to significant sources of bias (Villeneuve and Goldberg 2020). Compared with
56 studies on air pollution associated with the COVID-19 pandemic, more research has been
57 conducted on the correlations between temperature and COVID-19 transmission. Data-related
58 and methodological concerns are particularly prominent in the latter studies, inhibiting their
59 efforts to explicitly elucidate the complexity of the role of temperature in COVID-19 spread.
60 In this study, we first identified relevant reports and then attempted to explore the adequacy of
61 data and methods used, rather than concluded that whether temperature could influence the
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
5
62 COVID-19 transmission or not.
63
64 2. Methods
65 To identify articles associated with temperature and COVID-19 spread, we searched
66 ScienceDirect (https://www.sciencedirect.com/search), PubMed
67 (https://pubmed.ncbi.nlm.nih.gov/), and Web of Science (www.webofknowledge.com) using
68 the search terms “COVID-19” or “SARS-CoV-2” and “temperature” and “association” through
69 September 21, 2020. After examination of the titles, abstracts, and full text, 70 studies remained,
70 as illustrated in Figure 1. Since we excluded papers without peer review, we did not use other
71 search engines to examine pre-printed literature posted on the Internet.
72
73 3. Results and Discussion
74 Research status. The details of the 70 retrieved articles, including their location, study
75 design, adopted model, study period, confounding variables, and main findings, are presented
76 in Supplementary Material Table S1. Approximately 35 reports indicated a negative
77 association between temperature and COVID-19 transmission (Table 1), whereas 9 studies
78 suggested a positive association. Some researchers demonstrated that such associations were
79 piecewise, or found no clear link between temperature and COVID-19 spread. Regarding
80 location, approximately 73% of the studies (10 in one city and 41 in multiple regions) had been
81 conducted within one country. Of these 51 studies, 15 had been conducted in China; this is
82 unsurprising, because COVID-19 was first detected in Wuhan, China. Seven studies had been
83 conducted in the U.S. and India, followed by four in Spain, three in Brazil, and three in Japan
84 (Figure 2). 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
6
85 COVID-19 infections. As shown in Table 1, the daily new or cumulative COVID-19
86 counts were the most commonly adopted dependent variables, of which most were from official
87 health departments. During the early stage of the COVID-19 outbreak, the underreporting of
88 COVID-19 infections and deaths due to the lack of adequate testing in most countries might
89 have influenced the determined temperature-associated effects (Chatterjee 2020). Furthermore,
90 testing ability commonly increases as a pandemic evolves (Tromberg et al. 2020), thereby
91 inducing bias in the time-series analysis. Nonetheless, few of the reports we retrieved
92 considered the effects of testing ability in their analyses (Pan et al. 2020).
93 There are marked discrepancies in testing ability between regions worldwide
94 (
https://ourworldindata.org/coronavirus-testing#testing-for-covid-19-background-the-our-95 world-in-data-covid-19-testing-dataset). Testing coverage is particularly low in some
96 developing countries. Such inequalities should be inspected carefully because they may cause
97 considerable estimation errors in ecological studies (Iqbal et al. 2020; Pan et al. 2020). In
98 addition, uncertainties associated with asymptomatic COVID-19 infections or variations in
99 silent transmission between regions can significantly modify the estimation of the associations
100 between temperature and COVID-19 spread (Jia et al. 2020).
101 The changing definitions or misclassification of COVID-19 during the pandemic also
102 affected the COVID-19 counts. Using China as an example, the case definition was initially
103 narrow and was broadened later to include more infection cases as knowledge increased (Tsang
104 et al. 2020). However, most authors did not consider the effects of changing the case definition
105 in their statistical analyses.
106 Study design. Of the identified 70 publications, there are 24 ecological studies and 45 time
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
7
107 series studies (Table 1). Particularly, the time series studies can be further divided into two
108 types: temporal (31) and spatio-temporal (14) studies. Each study type has inherent possible
109 biases (Villeneuve and Goldberg 2020), i.e., the ecological fallacy or cross-level bias in the
110 ecological studies. The study design is particularly crucial in relation to the statistical models
111 and confounding variables. For example, in most temporal studies, the correlation analysis was
112 commonly adopted, without any confounding variables. Both the ecological and time series
113 studies can be analyzed by regression and correlation analysis. Some statistical models,
114 including the (S)ARIMA approach, are widely used in time series analysis.
115 Statistical model. Correlation analysis was conducted in more than 30% of the reports. In
116 particular, of the 21 studies that used correlation analysis, 13 implied a negative association,
117 whereas 9 exhibited a positive association (Table 1). The conclusions of the correlation
118 analyses were not always solid because they did not control for any other confounding factors,
119 which might have masked the true effect. Over the last 6 months, the temperature has increased
120 or decreased owing to seasonal changes. Meanwhile, the spread of COVID-19 has in some
121 cases been strongly suppressed by strict policy interventions (Lin et al. 2020). Thus, although
122 most of the reviewed authors declared that their correlation analysis results did not indicate
123 causality, these publications may still confuse public opinion regarding driving factors.
124 Regression models were also widely used in the retrieved studies. Most of the researchers
125 had conducted time-series analysis, whereas some did not follow the accepted methods of time
126 series analysis. We noted that multiple linear analysis was utilized in some studies (Haque and
127 Rahman 2020; Ladha et al. 2020), implying that the error in daily new cases was assumed to
128 have a normal distribution. For count data (such as infection cases), negative binomial, Poisson,
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
8
129 and zero-inflation regression models are more suitable to avoid overdispersion (Villeneuve and
130 Goldberg 2020).
131 Besides correlation and regression analyses, some of the researchers used machine learning
132 techniques(Malki et al. 2020; Pramanik et al. 2020). However, we found the methodologies of
133 these studies are not easy to follow (Malki et al. 2020; Pramanik et al. 2020), and their
134 conclusions evinced insufficient understanding of the mechanisms involved.
135 The factor of temperature. Another concern is how to choose a sound factor to represent
136 temperature. In the reviewed studies, the authors used the maximum, average, or minimum
137 daily temperature (Goswami et al. 2020), diurnal temperature range, moving average (Qi et al.
138 2020a; Xie and Zhu 2020), lagged effect (Briz-Redón and Serrano-Aroca 2020) and cross-basis
139 of temperature (Runkle et al. 2020; Shi et al. 2020), and yearly or monthly average temperature
140 (Mandal and Panwar 2020; Wei et al. 2020). However, at this stage, the differences in model
141 performance between these approaches remain unclear. Furthermore, as a large proportion of
142 the publications did not include sensitivity analysis or explain the reasons for their choices, we
143 cannot determine whether these choices were based on statistical significance, scientific
144 evidence, or other factors. In addition, the median incubation period for COVID-19 is estimated
145 to be 4–5 days, and incubation can extend to 14 days (Bi et al. 2020). Together with the
146 additional days for laboratory confirmation, using the temperature on the day of case
147 confirmation is not appropriate.
148 Meteorological factors and air pollutants. Approximately 25 studies did not include any
149 confounding variables, and most of these studies adopted correlation analyses. Most
150 confounding variables can fall into two types: the time-varying factors (meteorological factors,
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
9
151 air pollutants, policy intervention, and others) and location-varying factors (e.g., demography,
152 socioeconomic status, and population). Of the identified 70 studies, different confounding
153 factors pose threats to different types of studies. In particular, time-varying risk factors are
154 threats to both types of time series studies, whereas location-dependent factors are threats to
155 ecological and spatio-temporal but not purely temporal time series studies. With respect to
156 time-varying factors, we noted that approximately half of the retrieved reports controlled for
157 meteorological factors, particularly humidity, wind speed, and visibility (Table 1). However,
158 similar to the measurement for temperature, the lagged effects of meteorological factors should
159 be considered. Some studies conducted at the country or global scale just averaged the
160 temperature, the humidity, or other meteorological factors (Kumar and Kumar 2020; Sarmadi
161 et al. 2020), even though the weather conditions in some countries, such as the U.S., Russia,
162 India, and China, vary considerably. In contrast, the authors incorporated regional measures
163 for nationwide COVID-19 counts (Iqbal et al. 2020; Kumar and Kumar 2020; Sarkodie and
164 Owusu 2020; Sarmadi et al. 2020), because COVID-19 is prone to outbreaks in mega-cities,
165 particularly with more people traveling to and from international locations (E Dong et al. 2020).
166 Thus, appropriately weighting the corresponding meteorological factors between regions is
167 crucial to disentangle the temperature-related correlations.
168 Some of the retrieved studies also used air pollutants as covariates, such as particulate
169 matter, sulfur dioxide, and nitrogen dioxide (NO2) (Adhikari and Yin 2020; Azuma et al. 2020;
170 Jiang et al. 2020). The major objective of these studies was to explore the correlations between
171 exposure to air pollutants and COVID-19 transmission, considering air pollutants are widely
172 associated to human health (Wang et al. 2020). Some scientists have argued that such analyses
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
10
173 add incremental value during an active pandemic (Heederik et al. 2020; Villeneuve and
174 Goldberg 2020).
175 Policy interventions. Prior studies have demonstrated that strong policy interventions,
176 including face masks, social distancing, hand hygiene, travel or work restrictions, and
177 community isolation, can greatly lower the transmission of COVID-19 (Chu et al. 2020; Zhou
178 et al. 2020). However, only four of the retrieved studies controlled for social distancing (Li et
179 al. 2020; Rubin et al. 2020), non-pharmaceutical interventions (Fang et al. 2020), or strict
180 COVID-19 measures (Ozyigit 2020) in their analyses. In a time series analysis, policy
181 intervention would bend the growth curve in the later period of COVID-19 spread and also
182 decrease the reproduction number or prevent the number of positive counts (Davies et al. 2020).
183 It is questionable whether robust conclusions can be generated by models that omit policy
184 interventions. Existing studies have already determined that the stringency indexes for
185 governments’ responses (e.g., social distancing, school closing, and public event cancellation)
186 vary substantially between regions (Ashraf 2020; Hale et al. 2020). This spatial inequality
187 could reshape the curve between temperature and COVID-19 spread. However, none of the
188 studies we reviewed evaluated how the effects of spatial variations in the responses of
189 governments influenced the associations between temperature and COVID-19, especially in
190 ecological studies.
191 Location-varying factors. Approximately 50% of the publications included the effects
192 from location-varying factors, such as demographic factors, socioeconomic factors (e.g., race,
193 occupation, education, income, age structure, number of hospital beds, and life expectancy),
194 and spatiotemporal factors (e.g., number of days since the first confirmed case), especially in
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
11
195 the ecological and spatio-temporal studies. These time fixed factors that vary over locations
196 may modify the association of COVID-19 with temperature in multi-location temporal studies.
197 Research has shown that the age structures of North Americans and Europeans increase their
198 vulnerability to COVID-19 mortality (Esteve et al. 2020), which may be attributable to the
199 relatively high proportions of older people in these regions. Positive correlations were also
200 demonstrated (Figure S1) between the proportion of older people, testing number, life
201 expectancy, and gross domestic product per capita worldwide. Thus, researchers need to
202 carefully investigate the potential collinearities between the confounding variables before data
203 analysis. Some data processing techniques, such as principal component analysis and stratified
204 analysis, may be required prior to further analysis.
205 Study period and duration. Some ecological studies utilized the confirmed or
206 accumulative COVID-19 counts on a specific day as the dependent variable (Gupta et al. 2020;
207 Sarmadi et al. 2020). However, these COVID-19 data on a specific day may be greatly
208 influenced by the initial status, growth rate, and calendar date of the first case.
209 Furthermore, the exposure duration of more than 50% of the studies was in the range of
210 1–3 months or less than 1 month (Table 1). Some studies may only select a short study period
211 before the execution of policy intervention, and this short study period raises another issue: are
212 data from a short study period sufficient? Although there is no uniform criterion to determine
213 the minimum size for time-series studies, it is questionable whether a study period of 1-3
214 months is sufficient. For example, the determination of exposure to air pollution and mortality
215 generally requires a study period of multiple years to control for the long trend of adverse health
216 effects and address the seasonality of temperature (Z Dong et al. 2020; Yin et al. 2020).
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
12
217 To some extent, it is a paradox to researchers. At the early stage of pandemic, a number of
218 countries or regions were still in the stage of epidemic growth, and the growth curve may be
219 less influenced by policy intervention. However, an inherent question is the data that may be
220 not sufficient to account for temporal trend. Contrastingly, if longer study period is adopted,
221 associated parameters might be heavily determined by policy intervention, demographic
222 factors, and socioeconomic factors than by temperature.
223 Research areal unit. The authors of the retrieved studies investigated temperature and
224 COVID-19 transmission at the community, city, provincial or state, country, and global scales.
225 One study using the daily number of new cases nationwide in India revealed a positive
226 association (Kumar 2020), whereas provincial data in India suggested that temperature was
227 negatively associated with the number of COVID-19 cases (Goswami et al. 2020). This
228 difference may have been due to the modifiable area unit problem (MAUP), which is a form
229 of statistical bias that arises when incorporating point measurements into districts. A recent
230 study also found that the correlations between COVID-19 mortality and NO2 were
231 contradictory when aggregated at different levels, indicating that the MAUP should be
232 investigated when exploring the environmental determinants of the COVID-19 pandemic
233 (Wang and Di 2020).
234 Other issues. Other limitations were also noted. First, none of the existing studies
235 considered how the infectivity of the virus changed during the COVID-19 outbreak, although
236 this is an important time-varying factors. In addition, the geographical variations in the viral
237 strains with distinct infection capabilities may trigger biases in ecological studies. Second,
238 some of the authors adjusted the new/cumulative COVID-19 cases using the baseline on
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
13
239 previous days (Zhu and Xie 2020), whereas others did not (Qi et al. 2020b; Runkle et al. 2020).
240 Similarly, the population was not adopted as an offset in all of the studies (Qi et al. 2020b; Shi
241 et al. 2020). These variations in the data process may have hampered conclusions as to how
242 temperature affects the spread of COVID-19. Meanwhile, in some cases, COVID-19 infections
243 stem from clusters (for example, the worker in the food/meat processing industry or market)
244 rather than the whole population, which should be excluded or specified in statistical analysis.
245 Investigating the role of temperature in the COVID-19 pandemic is important but
246 challenging. Laboratory studies have observed that the high temperature may reduce the
247 survival of COVID-19 virus (Baker et al. 2020; NAS. 2020), while filed studies did not
248 consistently validate this conclusion. Our suggestion is that the study period should be taken
249 before the execution of policy intervention, since the policy intervention could strongly bend
250 the growth rate of COVID-19. In addition, comparing to ecological or time series studies, a
251 longitudinal study with individual data at global scale promises to better address the association
252 between temperature and COVID-19 transmission. Meanwhile, researchers also need to
253 carefully examine the influence from all potential confounding variables.
254 Also, we recommend that determining the influence of temperature on COVID-19
255 transmission can be comprehensively evaluated after the ending of this global pandemic. Till
256 now, the second wave of COVID-19 is still developing rapidly in some countries, implying
257 that temperature may be unable to significantly suppress COVID-19 transmission. A very
258 recent study concluded the weather contributed to 17% of the variation in the maximum
259 COVID-19 growth rate, and UV lights rather than temperature is the most strongly associated
260 with lower COVID-19 growth (Merow and Urban 2020). However, authors also pointed out
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
14
261 that the uncertainty remains high and aggressive policy interventions are likely be needed
262 (Merow and Urban 2020). Prior studies indicated that the variations of population susceptibility
263 is the driving factor of the COVID-19 pandemic, and warm temperature may be not anticipated
264 to substantially limit the COVID-19 growth (Baker et al. 2020; Su et al. 2020).
265
266 4. Conclusion
267 This study revealed that data-related and methodological issues mainly concerned data
268 reliability and processing, and the inherent uncertainties in the data decreased the reliability of
269 the statistical analyses. Since the COVID-19 pandemic begun, an enormous quantity of
270 manuscript submissions from the researchers in different countries or regions often led to the
271 need to perform the reviews in rush, which may be also responsible for some data and
272 methodological flaws, since many details might have been overlooked in these review
273 processes in order to provide the newest conclusions regarding the transmission and control of
274 COVID-19. From our point of view, most of the 70 peer-reviewed studies had significant flaws
275 in their methodologies or data design, requiring greater epidemiological rigor to yield robust
276 conclusions. Here we also encourage authors, reviewers, and editors to work together to more
277 closely scrutinize relevant research, aiming to produce studies with high-quality. With respect
278 to COVID-19 transmission, focusing more on the effectiveness and optimal range of
279 interventions, optimal strategies for reopening the economy and outdoor events, protective
280 materials, and tracing the sources of COVID-19 may be better assist in the global fight against
281 the COVID-19 pandemic.
282 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
15
283
5. Acknowledgements
284 This study was financially supported by a funding (Nos. GWTX05 and SWJC05) from the
285 National Institute of Environmental Health (NIEH), Chinese Center for Disease Control and
286 Prevention (China CDC). We thank Prof. Xiaoming Shi at NIEH, China CDC for his valuable
287 guidance and tremendous help for this study. We thank anonymous reviewers for their
288 insightful comments and constructive suggestions.
289
290
6. References
291 Adhikari A, Yin J. 2020. Short-term effects of ambient ozone, pm2. 5, and meteorological factors on COVID-19 292 confirmed cases and deaths in queens, new york. Int J Environ Res Public Health 17:4047.
293 Ashraf BN. 2020. Economic impact of government interventions during the COVID-19 pandemic: International 294 evidence from financial markets. Journal of Behavioral and Experimental Finance 27:100371.
295 Azuma K, Kagi N, Kim H, Hayashi M. 2020. Impact of climate and ambient air pollution on the epidemic growth 296 during COVID-19 outbreak in Japan. Environ Res:110042.
297 Baker RE, Yang W, Vecchi GA, Metcalf CJE, Grenfell BT. 2020. Susceptible supply limits the role of climate in 298 the early sars-cov-2 pandemic. Science 369:315-319.
299 Bi Q, Wu Y, Mei S, Ye C, Zou X, Zhang Z, et al. 2020. Epidemiology and transmission of COVID-19 in 391 300 cases and 1286 of their close contacts in shenzhen, China: A retrospective cohort study. The Lancet 301 Infectious Diseases.
302 Briz-Redón Á, Serrano-Aroca Á. 2020. A spatio-temporal analysis for exploring the effect of temperature on 303 COVID-19 early evolution in Spain. Sci Total Environ:138811.
304 Chatterjee P. 2020. Is india missing COVID-19 deaths? The Lancet 396:657.
305 Chu DK, Akl EA, Duda S, Solo K, Yaacoub S, Schünemann HJ, et al. 2020. Physical distancing, face masks, and 306 eye protection to prevent person-to-person transmission of sars-cov-2 and COVID-19: A systematic 307 review and meta-analysis. The Lancet.
308 Davies NG, Kucharski AJ, Eggo RM, Gimma A, Edmunds WJ, Jombart T, et al. 2020. Effects of non-309 pharmaceutical interventions on COVID-19 cases, deaths, and demand for hospital services in the uk: A 310 modelling study. The Lancet Public Health.
311 Dong E, Du H, Gardner L. 2020. An interactive web-based dashboard to track COVID-19 in real time. The Lancet 312 infectious diseases 20:533-534.
313 Dong Z, Wang H, Yin P, Wang L, Chen R, Fan W, et al. 2020. Time-weighted average of fine particulate matter 314 exposure and cause-specific mortality in China: A nationwide analysis. The Lancet Planetary Health 315 4:e343-e351.
316 Esteve A, Permanyer I, Boertien D, Vaupel JW. 2020. National age and co-residence patterns shape covid-19 317 vulnerability. PNAS 117:16118-16120.
318 Fang L-Q, Zhang H-Y, Zhao H, Che T-L, Zhang A-R, Liu M-J, et al. 2020. Meteorological conditions and 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
16
319 nonpharmaceutical interventions jointly determined local transmissibility of COVID-19 in 41 Chinese 320 cities: A retrospective observational study. The Lancet Regional Health-Western Pacific 2:100020. 321 Glasziou PP, Sanders S, Hoffmann T. 2020. Waste in COVID-19 research. BMJ.
322 Gong Y, Ma T-c, Xu Y-y, Yang R, Gao L-j, Wu S-h, et al. 2020. Early research on COVID-19: A bibliometric 323 analysis. The Innovation:100027.
324 Goswami K, Bharali S, Hazarika J. 2020. Projections for COVID-19 pandemic in india and effect of temperature 325 and humidity. Diabetes & Metabolic Syndrome: Clinical Research & Reviews.
326 Gupta A, Banerjee S, Das S. 2020. Significance of geographical factors to the COVID-19 outbreak in india. 327 Modeling Earth Systems and Environment:1-9.
328 Hale T, Petherick A, Phillips T, Webster S. 2020. Variation in government responses to COVID-19. Blavatnik 329 school of government working paper 31.
330 Haque SE, Rahman M. 2020. Association between temperature, humidity, and COVID-19 outbreaks in 331 bangladesh. Environ Sci Policy.
332 Heederik DJ, Smit LA, Vermeulen RC. 2020. Go slow to go fast: A plea for sustained scientific rigour in air 333 pollution research during the COVID-19 pandemic.Eur Respiratory Soc.
334 Iqbal MM, Abid I, Hussain S, Shahzad N, Waqas MS, Iqbal MJ. 2020. The effects of regional climatic condition 335 on the spread of COVID-19 at global scale. Sci Total Environ 739:140101.
336 Jia X, Chen J, Li L, Jia N, Jiangtulu B, Xue T, et al. 2020. Modeling the prevalence of asymptomatic COVID-19 337 infections in the Chinese mainland. The Innovation 1:100026.
338 Jiang Y, Wu X-J, Guan Y-J. 2020. Effect of ambient air pollutants and meteorological variables on COVID-19 339 incidence. Infect Control Hosp Epidemiol:1-11.
340 Jüni P, Rothenbühler M, Bobos P, Thorpe KE, da Costa BR, Fisman DN, et al. 2020. Impact of climate and public 341 health interventions on the COVID-19 pandemic: A prospective cohort study. CMAJ 192:566-573. 342 Kraemer MU, Yang C-H, Gutierrez B, Wu C-H, Klein B, Pigott DM, et al. 2020. The effect of human mobility 343 and control measures on the COVID-19 epidemic in China. Science 368:493-497.
344 Kumar G, Kumar R. 2020. A correlation study between meteorological parameters and COVID-19 pandemic in 345 mumbai, india. Diabetes & Metabolic Syndrome: Clinical Research & Reviews.
346 Kumar S. 2020. Effect of meteorological parameters on spread of COVID-19 in india and air quality during 347 lockdown. Sci Total Environ 745:141021.
348 Ladha N, Bhardwaj P, Charan J, Mitra P, Goyal JP, Sharma P, et al. 2020. Association of environmental 349 parameters with COVID-19 in delhi, india. Indian J Clin Biochem:1-5.
350 Li AY, Hannah TC, Durbin JR, Dreher N, McAuley FM, Marayati NF, et al. 2020. Multivariate analysis of black 351 race and environmental temperature on COVID-19 in the US. The American journal of the medical 352 sciences.
353 Lin S, Wei D, Sun Y, Chen K, Yang L, Liu B, et al. 2020. Region-specific air pollutants and meteorological 354 parameters influence COVID-19: A study from mainland China. Ecotoxicol Environ Saf 204:111035. 355 Malki Z, Atlam E-S, Hassanien AE, Dagnew G, Elhosseini MA, Gad I. 2020. Association between weather data 356 and COVID-19 pandemic predicting mortality rate: Machine learning approaches. Chaos, Solitons & 357 Fractals:110137.
358 Mandal CC, Panwar M. 2020. Can the summer temperatures reduce COVID-19 cases? Public Health 185:72-79. 359 Merow C, Urban MC. 2020. Seasonality and uncertainty in COVID-19 growth rates [just accepted]. Doi: 360 10.1073/pnas.2008590117. Proc Natl Acad Sci U S A.
361 NAS. 2020. Rapid expert consultation on sars-cov-2 survival in relation to temperature and humidity and potential 362 for seasonality for the COVID-19 pandemic. Nas: National academies of sciences engineering and 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
17
363 medicine., available:
Https://www.Nap.Edu/catalog/25771/rapid-expert-consultation-on-sars-cov-2-364
survival-in-relation-to-temperature-and-humidity-and-potential-for-seasonality-for-the-covid-19-365 pandemic-april-7-2020. Doi: Https://doi.Org/10.1016/s2666-5247(20)30003-3.
366 Ozyigit A. 2020. Understanding covid-19 transmission: The effect of temperature and health behavior on 367 transmission rates. Infection, Disease & Health.
368 Pan J, Yao Y, Liu Z, Meng X, Ji JS, Qiu Y, et al. 2020. Warmer weather unlikely to reduce the COVID-19 369 transmission: An ecological study in 202 locations in 8 countries. Sci Total Environ:142272.
370 Pramanik M, Udmale P, Bisht P, Chowdhury K, Szabo S, Pal I. 2020. Climatic factors influence the spread of 371 COVID-19 in russia. Int J Environ Health Res:1-15.
372 Qi H, Xiao S, Shi R, Ward MP, Chen Y, Tu W, et al. 2020a. COVID-19 transmission in mainland China is 373 associated with temperature and humidity: A time-series analysis. Sci Total Environ 728:138778. DOI: 374 138710.131016/j.scitotenv.132020.138778.
375 Qi H, Xiao S, Shi R, Ward MP, Chen Y, Tu W, et al. 2020b. COVID-19 transmission in mainland China is 376 associated with temperature and humidity: A time-series analysis. Sci Total Environ:138778.
377 Rubin D, Huang J, Fisher BT, Gasparrini A, Tam V, Song L, et al. 2020. Association of social distancing, 378 population density, and temperature with the instantaneous reproduction number of sars-cov-2 in 379 counties across the united states. JAMA network open 3:e2016099-e2016099.
380 Runkle JD, Sugg MM, Leeper RD, Rao Y, Mathews JL, Rennie JJ. 2020. Short-term effects of weather parameters 381 on COVID-19 morbidity in select US cities. Sci Total Environ:140093.
382 Sarkodie SA, Owusu PA. 2020. Impact of meteorological factors on COVID-19 pandemic: Evidence from top 20 383 countries with confirmed cases. Environ Res:110101.
384 Sarmadi M, Marufi N, Moghaddam VK. 2020. Association of COVID-19 global distribution and environmental 385 and demographic factors: An updated three-month study. Environ Res 188:109748.
386 Shi P, Dong Y, Yan H, Zhao C, Li X, Liu W, et al. 2020. Impact of temperature on the dynamics of the COVID-387 19 outbreak in China. Sci Total Environ:138890.
388 Su M, Peng S, Chen L, Wang B, Wang Y, Fan X, et al. 2020. A warm summer is unlikely to stop transmission of 389 COVID‐19 naturally. GeoHealth:e2020GH000292.
390 Tromberg BJ, Schwetz TA, Pérez-Stable EJ, Hodes RJ, Woychik RP, Bright RA, et al. 2020. Rapid scaling up of 391 covid-19 diagnostic testing in the united states—the nih radx initiative. N Engl J Med.
392 Tsang TK, Wu P, Lin Y, Lau EH, Leung GM, Cowling BJ. 2020. Effect of changing case definitions for COVID-393 19 on the epidemic curve and transmission parameters in mainland China: A modelling study. The Lancet 394 Public Health.
395 Villeneuve PJ, Goldberg MS. 2020. Methodological considerations for epidemiological studies of air pollution 396 and the sars and COVID-19 coronavirus outbreaks. Environ Health Perspect 128:095001.
397 Wang H, Lu F, Guo M, Fan W, Ji W, Dong Z. 2020. Associations between PM1 exposure and daily emergency 398 department visits in 19 hospitals, Beijing. Sci Total Environ 755:142507.
399 Wang Y, Di Q. 2020. Modifiable areal unit problem and environmental factors of COVID-19 outbreak. Sci Total 400 Environ:139984.
401 Wei J-T, Liu Y-X, Zhu Y-C, Qian J, Ye R-Z, Li C-Y, et al. 2020. Impacts of transportation and meteorological 402 factors on the transmission of COVID-19. Int J Hyg Environ Health:113610.
403 Xie J, Zhu Y. 2020. Association between ambient temperature and COVID-19 infection in 122 cities from China. 404 Sci Total Environ 724:138201.
405 Yin P, Guo J, Wang L, Fan W, Lu F, Guo M, et al. 2020. Higher risk of cardiovascular disease associated with 406 smaller size-fractioned particulate matter. Environmental Science & Technology Letters 7:95-101. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
18
407 Zeka A, Tobias A, Leonardi G, Bianchi F, Lauriola P, Crabbe H, et al. 2020. Responding to COVID-19 requires 408 strong epidemiological evidence of environmental and societal determining factors. The Lancet Planetary 409 Health 4:e375-e376.
410 Zhou Y, Xu R, Hu D, Yue Y, Li Q, Xia J. 2020. Effects of human mobility restrictions on the spread of COVID-19 411 in shenzhen, China: A modelling study using mobile phone data. The Lancet Digital Health 2:e417-e424. 412 Zhu Y, Xie J. 2020. Association between ambient temperature and COVID-19 infection in 122 cities from China. 413 Sci Total Environ 724:138201. DOI: 138210.131016/j.scitotenv.132020.138201.
414 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
19
415 Table 1. Summary of the number of peer-reviewed publications estimating associations between temperature and COVID-19 transmission.
Main findings of available literature
Items Categories All Negative
association Positive association Piecewise correlatio n No association Others
Within one city 10 4 3 0 3 0
Within multiple cities in one country 41 19 7 5 4 6
Study location Multiple countries 19 12 2 0 2 3 <1 month 19 11 4 0 2 2 1–3 months 37 18 5 5 6 3 Study perioda >3 months 10 4 3 0 1 2 Ecological 24 13 4 1 4 2
Time series: temporal association 31 14 8 1 3 5
Time series: spatio-temporal 14 8 0 3 2 1
Study design Descriptive 1 0 0 0 0 1 Correlation analysis 21 13 7 0 0 1 Regression analysis 45 20 6 5 7 7 Methodologyb Other approaches 12 6 0 0 3 3
No confounding factor considered 25 12 7 1 2 0
Meteorological factors 33 14 4 3 6 0
Air pollutants 7 4 2 0 1 0
Interventions 4 2 0 1 0 0
Confounding variablesb
Demographic and socioeconomic variables and spatial and temporal effects 35 18 4 3 6 0 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
20
Daily new counts 31 15 7 4 4 5
Daily accumulative counts 10 7 2 1 3 1
New/accumulative counts on a specific
day 9 5 2 0 0 2
Reproduction number 9 3 0 2 3 1
Dependent variablesc
Others 11 7 2 0 1 1
All 70 35 12 5 9 9
416 Note: a, some reports did not have detailed information on the study period; b, some reports used multiple methodologies or confounding variables;
417 and c, some publications used multiple dependent variable 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
21
418
Article identified using keywords (n=1304)
Sciencedirect (n=1071)
PubMed (n=160)
Web of Science (n=73)
Duplicate removed (n=136)
Articles identified for further analysis (n=1168)
Removed after the examination on the title and abstract (n=1062)
Articles identified for further analysis (n=106)
Removed after the examination on the full-text (n=36)
Articles included in presented analysis (n=70)
419 Figure 1. The process for the peer-reviewed publication identification in the present
420 study. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Accepted Manuscript
22
421
422 Figure 2. The global distribution for the number of peer-reviewed publications on the
423 association between temperature and COVID-19 spread (data as of September 14,
424 2020). 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
